Using Telephone Reminders to Increase Attendance at Psychiatric Appointments: Findings of a Pilot Study in Rural Appalachia

Patients do not receive the mental health treatment they need when they miss appointments. Missed appointments also result in inefficient use of agency resources. Some research findings suggest that phone reminders increase attendance at initial appointments ( 1 , 2 , 3 , 4 ), ongoing therapy appointments ( 5 ), and aftercare appointments ( 6 ). Findings from other studies lend doubt about the effectiveness and efficiency of phone reminders in increasing kept appointments.

Stasiewicz and Stalker ( 7 ) randomly assigned 128 individuals seeking substance abuse treatment to one of four conditions: intake appointment scheduled within 48 hours of the client's call, intake appointment scheduled 48 hours or more after the call and a phone reminder the day before the intake appointment, intake appointment scheduled 48 hours or more after the call and a mailed appointment card and clinic brochure, or intake scheduled 48 hours or more after the client's call. Approximately 72% of the patients scheduled within 48 hours kept their appointments, significantly higher than the rates (50% to 53%) in the other groups. In another study, phone reminders for initial appointments significantly increased appointment keeping only when the appointments were scheduled within a week of the patient's call for an appointment ( 8 ). Shivack and Sullivan ( 9 ) found that phone reminders for first appointments resulted in significantly higher attendance than no reminder calls only when patients were reached directly.

This brief report describes a pilot study that examined whether reminder calls increased attendance at intake and therapy appointments at a mental health center in rural Appalachia. It also examined whether there were differences in appointment keeping based on whether the therapist or a staff member placed the call. Subsequent to findings from the pilot study, a ten-week trial period ensued in which all therapists were asked to make reminder calls to patients scheduled for intake appointments. Findings from the trial period are also reported.

Methods

The first pilot study was conducted in two rural clinics, approximately 80 miles apart, of a community mental health center in rural Appalachia. Patients travel up to 30 miles for appointments. The agency's cancellation and no-show rate ranges from 21% to 26% for therapy appointments and from 22% to 43% for intake appointments (Aiken J, Buuck J, Charlton P, et al, unpublished manuscript, 1993). Staff often phone patients to remind them of appointments, but this is not consistently done.

The study was conducted from the end of June 2005 to the end of August 2005. One of three phone call conditions was assigned to each of the nine weeks. The conditions were no reminder calls, calls by the therapist, and calls by a staff member. Each condition was written on a slip of paper, and one slip was drawn and assigned to one of the clinics for the first week. One of the two remaining slips was drawn and assigned to the second clinic for the first week. Both slips were returned to the "hat," and the procedure was followed for the remaining eight weeks. Two restrictions were applied to the assignment of conditions. First, the no-call condition was not assigned to two sequential weeks in either clinic. Second, because there were three conditions and nine weeks, each condition was applied for three weeks in each clinic.

One therapist at each clinic participated in the study, and only their patients were included in the study. This is a major limitation; however, the findings had the potential to raise important practice and research questions. Moreover, the study findings resulted in a second pilot study in which all agency therapists were expected to phone and remind patients of their intake appointments. This was instituted from April 7, 2006, to June 13, 2006. For both pilot studies phone reminders were made the day before the appointment. Calls were made on Friday for Monday appointments. Staff reminded patients of their appointments, whereas therapists introduced themselves (for intake appointments) as the therapist who would be seeing the patient.

The following information was recorded for each scheduled appointment: whether a phone call was attempted, who made the call, outcome of the call (talked to patient, left message on answering machine, left message with family or friend, or no answer or no phone), and whether the appointment was kept. Information was obtained on whether the appointment was for an intake or therapy appointment, whether it was for an adult or a child, and whether the patient called to cancel the appointment. The study was approved by the institutional review board at Ridgeview Psychiatric Hospital and Center.

Some patients had multiple appointments that they could either keep or not keep, and each therapist saw some patients more than once. Therefore, the analysis strategy needed to address nonindependence of the observations. MIXOR (mixed-effects ordinal regression) is a computer program developed by Hedeker and Gibbons for the analysis of data from clustered and longitudinal designs ( 10 ). MIXOR can be used with dichotomous and ordinal response-level observations and takes into account dependency in the data. The use of mixed-effects logistic regression with MIXOR overcame the data dependence problem, because the model does not assume the observations are independent; the model parameters are adjusted on the basis of the degree of dependency in the data.

The independent variables were clinic site, age of patient (adult or child), and four dummy variables representing phone call condition. The five phone call conditions were no contact, therapist directly contacted patient, therapist indirectly contacted patient, staff directly contacted patient, and staff indirectly contacted patient. The reference group was the control condition of no contact. Interactions between each of the dummy variables and type of appointment (intake or therapy) were also tested.

Results

There were 451 appointments scheduled across the nine weeks. A total of 310 (69%) were therapy appointments, and 141 (31%) were intake appointments. Fourteen (10%) of the intake appointments were cancelled by patients. The cancellation rate for therapy appointments was 18% (N=56). Cancelled appointments were not included in the analysis. The final sample consisted of 381 appointments (254 therapy and 127 intake) for 238 patients. The number of therapy appointments per patient ranged from one to nine. Most patients had only one scheduled intake appointment, although three patients had two scheduled intake appointments. Patients could have more than one scheduled intake appointment if the first one was not kept. Two patients kept neither of the intake appointments, and one patient kept the second scheduled intake appointment. Eight patients had both intake and therapy appointments during the nine-week period.

Table 1 shows the number of scheduled intake and therapy appointments in each phone contact condition. The no-contact condition consisted of appointments not scheduled for a reminder call and appointments scheduled for a call in which the patient had no phone or had a phone but could not be reached (no answer or no answering machine). Indirect contact was made when a message was left on the answering machine or with friends or family. Direct contact was made when the caller talked with the patient to remind him or her of the appointment.

Enlarge table

Seven independent variables (clinic site, age of patient, type of appointment, and four dummy variables representing phone call condition) were entered into a mixed-effects logistic regression model. (Reported probability levels are two-tailed). There was a significant effect for type of appointment (z=4.44, p<.001). As shown in Table 2 , of the 254 therapy appointments, 210 appointments (83%) were kept, compared with 69 of the 127 intake appointments (54%). Compared with patients who were not contacted, patients who were reached directly by their therapists were significantly more likely to keep their appointments (z=3.03, p=.002). The other phone call conditions had no significant effect on appointment keeping. A significant interaction was found between direct contact by the therapist and type of appointment (z=1.95, p=.05).

Enlarge table

Compared with no contact, therapist direct contact resulted in an increased rate of appointment keeping. However, the rate of increase depended on the type of appointment. The rate of appointment keeping for therapy when the therapist directly contacted patients was 89%, compared with 79% when there was no contact. Only 43% of the intake appointments were kept when there was no contact, compared with 94% when the therapist reached the patient directly. For intake appointments, but not for therapy appointments, the rate of appointment keeping decreased sharply when patients were not contacted directly by their therapists. Compared with intake appointments, therapy appointments depended less on the phone call condition for the rate of appointment keeping. The range of keeping therapy appointments was 79% (no contact) to 89% (therapist direct contact) across the five phone contact conditions, whereas the rate for keeping intake appointments ranged from 43% (no contact) to 94% (therapist direct contact). For the other three conditions, the rate of intake appointment keeping was 65%, 55%, and 46%, for therapist indirect contact, staff direct contact, and staff indirect contact, respectively. The rate of therapy appointment keeping was 83%, 84%, and 82%, for therapist indirect contact, staff direct contact, and staff indirect contact, respectively.

As a result of this practice-based pilot study, the agency decided to conduct a second small project. From April 7, 2006, to June 13, 2006, all therapists were asked to call and remind patients who were scheduled for intake appointments. This involved ten therapists across four clinics. There were 311 scheduled intake appointments during this period. Patients rescheduled, cancelled, or called in sick for 28 of the intakes, a cancellation rate of 9%. Excluding the cancelled appointments, there were 283 intake appointments, of which 146 (52%) were kept and 137 (48%) were not kept. The 283 intake appointments were scheduled for 259 patients. Most patients had only one scheduled intake appointment, but some had two or three intakes scheduled.

Direct contact was made for 29% (N=82) of the scheduled intake appointments, indirect contact was made for 29% (N=83) of the appointments, and for 42% (N=118) of appointments no contact was made. Patients who were directly contacted kept 74% of their appointments (N=61), patients who were indirectly contacted kept 49% of their appointments (N=41), and patients who were not contacted kept 37% of their appointments (N=44).

The findings reinforce those from the first pilot study. Direct contact was statistically significant and resulted in more kept appointments than no contact (z=2.03, p=.02). Indirect contact, compared with no contact, did not result in significantly more appointment keeping.

Discussion

A study limitation of the first pilot study is that it was based in two clinics and included only two therapists. The initial project was kept small because of concern that not calling patients would result in their not keeping appointments and thus not receiving services. However, the second pilot study was implemented in response to these limitations, and the findings replicated those from the first study. We also did not control for the amount of time patients had to wait for their intake appointments, but the waiting time across clinics was approximately two to three weeks. Clinic site was included as an independent variable and was not significantly associated with appointment keeping. This, as well as the similarity in waiting time across clinics, mitigated the limitation of not including wait time as a variable. The findings extend those from other studies in that we examined whether who made the reminder call and appointment type affected appointment keeping.

The findings support those of Shivack and Sullivan ( 9 ) who found that reminder calls were most effective when patients were reached directly. Our findings also replicated those from a study that showed that any verbal contact with the assigned intake therapist before the initial appointment significantly increased the rate of kept appointments ( 11 ). Other studies have not examined whether who makes the reminder call makes a difference, but together these findings suggest that it does, at least for intake appointments. An explanation for this is that early contact with the therapist begins to build the therapeutic relationship and reassures patients that the therapist is a caring person. However, this rationale needs to be subjected to empirical testing. For now, agencies should consider giving therapists time to place reminder calls for intake appointments, monitor whether doing so makes a difference in appointment keeping, and report the results so the data can inform practice.

In the first pilot study, only 8% (N=23) of those scheduled for a call had no phone. The problem remains of how to directly reach patients who do have a phone. Of the patients with therapy appointments scheduled for a reminder call, 28% (N=52) were left a message and 24% (N=46) could not be reached at all. Of the intake appointments scheduled for a call, 30% (N=30) were left a message and 33% (N=33) could not be reached at all. In the second study, only 35% of the patients were reached directly as a result of a call attempt. We did not keep track of the number of calls placed before reaching the patient, but this would be useful in assessing the efficiency of using reminder calls.

Conclusions

Reminder calls made little difference in appointment keeping, except for intake appointments when patients were contacted directly by their therapists. Several implications emerge from the findings. First, close to 80% of therapy appointments were kept when patients were not reminded of their appointments. It may not be efficient to make reminder calls for therapy appointments. On the other hand, patients who miss appointments because they do not receive a reminder call will not receive treatment. Furthermore, because of the finding that a direct therapist contact makes a difference for intake appointments, it may also make a difference for the first few therapy appointments. We did not track which therapy appointment it was, but it is important to do so in future studies and to sort out whether phone reminders are more or less effective at different stages of therapy. Second, agencies may want to provide time for therapists to make reminder calls for initial appointments. Third, the findings need to be replicated with larger samples and in other settings. More research is needed to understand which factors interfere with appointment keeping and which agency characteristics and clinician behaviors are associated with patients' keeping appointments. For example, the length of time between the intake call and the first appointment may moderate the impact of phone call reminders on appointment keeping.

Phone reminders, reduced waiting time, and other strategies all deserve more research on their effectiveness, alone or in combination, to increase appointment keeping. It would be helpful to learn more about why patients do not keep appointments and how they view reminder calls. Given the high rate of missed appointments across agencies and client groups, it may be that a simple intervention like phone reminders can increase the rate of kept appointments, especially for initial appointments.

Acknowledgments and disclosures

This research was supported by grant 5K01-MH-068473 from the National Institute of Mental Health to Dr. Staudt.

The authors report no competing interests.